HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by OLESKE, D. M. Articles by SMITH, T. Search for Related Content PubMed PubMed Citation Articles by OLESKE, D. M. Articles by SMITH, T.

Effect of Medicaid Managed Care on Pregnancy Complications

From the Department of Health Systems Management, Department of Preventive Medicine, Rush University, Chicago, Illinois; Advocate Lutheran General Hospital, Park Ridge, Illinois; the Agency for Health Care Policy and Research, Rockville, Maryland; and the Maternal and Child Health Branch, California Department of Health Services, Sacramento, California.

Address reprint requests to: Denise M. Oleske, PhD, Department of Health Systems Management, Rush University, 1653 West Congress Parkway, Chicago, IL 60612

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Objective: This study examined the impact of managed care on hospital obstetric outcomes in Medicaid-sponsored women.

Methods: The study sample consisted of a total of 525,517 maternal deliveries for singleton births from three payer groups, Medicaid managed care, Medicaid fee-for-service, and private managed care in 439 short-term-stay nonfederal hospitals in California and Florida. Quality of care comparisons were made using six indicators. Data were derived from linked computer files of birth certificates, hospital discharge abstracts, Medicaid eligibility records, Medicaid health care claims, and surveys of hospital characteristics.

Results: The overall multivariate likelihood of an adverse maternal outcome during hospitalization for a delivery was not significantly different between Medicaid managed care and Medicaid fee-for-service groups in California and Florida. However, mothers in the Medicaid managed care group compared with mothers in the private managed care group experienced a higher likelihood of eclampsia (California) (adjusted odds ratio = 1.26; 95% confidence interval 1.05, 1.57; P = .04).

Conclusion: Overall, managed care has not adversely affected pregnancy outcomes in Medicaid-sponsored women. Yet, payer system changes may be insufficient to achieve complete parity of outcomes relative to private managed care patients.

During the period 1985 to 1993, the number of beneficiaries in the Medicaid program increased by 50% and Medicaid costs tripled.¹ In 1993, Medicaid was a $131 billion health care program for 33 million low-income Americans. Expecting the trends to continue to increase, states began channeling Medicaid-eligible individuals into managed care plans in an attempt to control health care costs and improve access.² However, the impact of this transition on the quality of care for poor and underserved women had not been extensively examined in terms of the outcomes of care provided at the time this study began. This is of particular concern given the increasing numbers of women who are uninsured or underinsured at the time of their delivery. When providing care to Medicaid patients, hospitals are at financial risk. Hospitals billing Medicaid on a per diem basis may incur financial losses if serving a large proportion of high-risk pregnant women who require many services. Hospitals in risk contracts with a state’s Medicaid agency receive a global reimbursement based on a patient’s diagnostic category, and thus may experience financial losses by not receiving adequate reimbursement for the additional days of hospital stay often required by these medically high-risk mothers. Faced with decreasing monetary resources, hospitals must change modes of service delivery to respond to these environmental pressures. If changes in hospital payment result in hospital services being compromised to a vulnerable group of women, will this lead to an increased incidence of adverse maternal outcomes?

Although a large literature exists on the relationship of Medicaid payment for prenatal care and maternal and newborn outcomes, only three studies using small samples have investigated the impact of managed care on outcomes of prenatal care.^3–5 None have compared Medicaid managed care and Medicaid fee-for-service for hospital obstetric care with respect to pregnancy outcomes. Instead, studies to date of the relationship between expected primary payer and quality of hospital obstetric care have used either the cesarean delivery rate or the vaginal birth after cesarean (VBAC) rate as indicators of quality.^3,6–9 Unfortunately, most of the previous studies comparing these rates do not distinguish between those covered by Medicaid fee-for-service from those covered by Medicaid managed care groups. The controversy over the significance and interpretation of these rate measures of quality points to the need for a broad range of indicators to assess differences in the quality of hospital obstetric care between Medicaid fee-for-service and Medicaid managed care, as no singular perfect measure of hospital obstetric quality exists. The controversy also emphasizes the lack of outcome measures for the most common reason for hospitalization in the United States: namely, obstetric delivery.

The purpose of this study was to investigate the impact of managed care in the Medicaid population by comparing the maternal outcomes of those in fee-for-service arrangements with those in managed care programs. In addition, maternal outcomes in the Medicaid managed care group were compared with those in private managed care to determine whether managed care would differently affect a vulnerable population such as that covered by Medicaid. This study may be viewed as a baseline for the continued monitoring of the effects of the transition to increasing enrollment of Medicaid-eligible individuals in managed care plans on the health outcomes for women with newborn deliveries.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
The study population consisted of singleton newborn deliveries in short-stay nonfederal hospitals in 1993 in California and Florida. Newborn deliveries were selected from three payer groups: Medicaid managed care, Medicaid fee-for-service, and private managed care. Singleton newborn deliveries were identified from the birth certificate and from the hospital discharge abstracts diagnosis codes V30.00 or V30.01 or as admission status of "newborn" on the hospital discharge abstract. Births weighing less than 500 g were excluded from the sample, as biologic factors are more likely to affect outcome than the health care system. Maternal deliveries were identified from the hospital discharge abstract computer file if the record was coded as vaginal or cesarean delivery (any Diagnostic Related Group numbers 370 through 375, inclusive) and the mother’s age at delivery was between 10 and 50 years. Hospitals eligible for study had to have at least ten births in the year of study.

In the study year, the type and extent of managed care in the study states was as follows. In California, 680,043 of California’s Medicaid recipients were enrolled in a managed care plan. The managed care organizational types used for delivering care to the Medicaid population in California were: ten health maintenance organizations, three primary care case management plans, 21 prepaid health plans, and two health-insuring organizations (in Santa Barbara and San Mateo counties). None of these plans was statewide, as each of the plans served only Medicaid beneficiaries residing within specific geographic boundaries, usually based on county lines. (The Medicaid program in California is called Medi-Cal, but for purposes of uniformity, the term "Medicaid" is used for both California and Florida.) In Florida, 16 health maintenance organizations participated in the Medicaid managed care plans, all of which were using full capitation. One additional primary care case management plan was using the fee-for-service arrangement for reimbursement. Total Medicaid enrollment in these plans in 1993 was 382,978. Only two of 17 plans were statewide plans. The majority of health maintenance organizations participating in the Medicaid program were located in south Florida. Despite heterogenity of organizational forms of managed care in the study states, all plans coded as "managed care" involved some degree of capitation.

The determination payer group classification differed because of differences in the availability of payer information on the various data sources used in the study states. In California, the Medicaid eligibility files were used for assignment of a record into payer group Medicaid managed care or Medicaid fee-for-service. In Florida, a record was assigned to Medicaid managed care, Medicaid fee-for-service, or private managed care based on the expected payer code recorded on the hospital discharge abstract.

Six maternal outcomes were compared in the payer groups which were thought to be quality indicators of hospital obstetric care: 1) in-hospital death, 2) eclampsia/convulsions, 3) excessive blood loss as demonstrated by transfusion requirement, 4) anesthesia complications, 5) cardiac arrest, and 6) trauma to perineum or vulva. The maternal outcomes selected for study were proposed as quality indicators by the ACOG and the Joint Commission on Accreditation of Healthcare Organizations.^10,11 The indicators selected for study were known to have high sensitivity in large databases and were believed to be preventable during, or influenced by, the hospital stay.^12–15 Information from the computer files of birth certificates as well as the hospital discharge abstracts were searched to determine whether an outcome was present.

Maternal confounding variables measured were: age at delivery (in years), race (black compared with non-black), ethnicity (Hispanic compared with non-Hispanic), parity (one or more compared with none), educational level (less than high school compared with high school or more), delivery type (cesarean compared with vaginal), adequacy of prenatal care (inadequate or adequate/adequate plus), and high-risk status (yes or no). The Kotelchuk index measured the adequacy of prenatal care.^16,17 The cutpoint from the index used for adequate prenatal care used was based on the ACOG recommendation that adequate prenatal care is initiated during the first 4 months of pregnancy followed by at least 80% of the total number of visits recommended by ACOG.¹⁸ Maternal demographic characteristics and prenatal care were derived from the birth certificate. High-risk status for obstetric complications was defined according to the Guidelines for Perinatal Care¹⁹ using the International Classification of Diseases, 9th Revision (ICD-9) diagnoses codes on the maternal hospital discharge abstract. Newborn birth weight was obtained from the birth certificate. Extensive quality control procedures are undertaken by California and Florida to ensure the accuracy of data entered into computer tapes of birth certificates and hospital discharge abstracts.²⁰ Hospital confounding variables included in multivariate analyses were: teaching status (Council of Teaching Hospitals member, medical school affiliation, or presence of residency programs) and presence of selected continuous quality process initiatives in hospital obstetric care. Information on hospital characteristics was obtained from the American Hospital Association’s Annual Survey of Hospitals²¹ and from a survey we administered to hospitals in the study states.

The strategy for combining patient level data from the computer files focused on data elements common to the data sets. Mothers’ hospital discharge abstracts and birth certificates were linked using the hospital identification number, mother’s date of birth, admission and discharge date for the delivery, mother’s race/ethnicity, ZIP code of residence, and delivery method. Hospital characteristics were merged with the linked file of mother and newborn characteristics through an intermediate file of hospitals using the American Hospital Association–assigned identification number. Linkage rates of 89.5% for California and 93.4% for Florida were achieved. Because confidential patient level information was used for the linkage processes, the institutional review boards for the Vital Statistics Offices and the government offices which managed the hospital discharge abstract files of the States of California and Florida as well as that of the first author’s medical center reviewed and approved this research study before its implementation.

To assess the associations of payer and maternal pregnancy outcomes, multiple logistic regression analyses were performed. The analytic strategy was aimed at answering the primary research question, "Can managed care as an alternative payer system provide the same level of quality for hospital obstetric care as fee-for-service among Medicaid women?" Hence, the payer contrasts were between the Medicaid managed care and Medicaid fee-for-service groups, and between the Medicaid managed care and the private managed care groups (women in the same payer system). The two study states were modeled separately to provide an opportunity to see whether state-level experience with managed care would influence the results, with California having far greater experience in the delivery of health care through managed care to its Medicaid population.

Analysis of the likelihood of maternal adverse outcomes (no = 0, yes = 1) from multiple logistic regression included payer-pair contrast (Medicaid managed care = 1, Medicaid fee-for-service = 0, and private managed care = 0) and the following risk factors a priori specified (1 = factor present, 0 = factor absent): black race, Hispanic ethnicity, high parity, low educational level, cesarean delivery, less than adequate prenatal care, low birth weight, high intensity of hospital obstetric quality initiatives, delivery in teaching hospital, and maternal high-risk status. Maternal age at delivery was a continuous variable. In addition, interaction terms consisting of payer–race and payer–ethnicity were included in all models because of significant associations detected at the bivariate level of race and ethnicity with various outcomes. The variable high-risk status was excluded in models of the adverse outcome, eclampsia.

The individual maternal adverse outcomes were examined both separately and combined for analyses. Each state was also modeled separately to contrast each state’s experience level with managed care. California’s experience with managed care began in the early 1980s, whereas in Florida managed care was not introduced until the early 1990s.

	Results

Top Abstract Materials and Methods Results Discussion References

 
In California, the Medicaid managed care group had the largest proportion of women who were black, were Hispanic, did not complete high school or were under 20 years of age (P < .001). The private managed care group had a significantly larger percentage of women at high risk for obstetric complications than the Medicaid managed care group (P < .001), but the Medicaid managed care and the Medicaid fee-for-service groups were similar with respect to the percentage of women classified as high risk. The percentage of women with adequate prenatal care in the Medicaid managed care group was similar to the Medicaid fee-for-service group, but significantly lower than the private managed care group (53.64% compared with 79.28%, P < .001) (Table 1).

	Discussion

Top Abstract Materials and Methods Results Discussion References

In analyzing the impact of payer system on pregnancy obstetric outcomes, we controlled for maternal (age, race, education, parity, prenatal, and clinical conditions), prenatal care, and hospital (teaching status and quality initiatives) characteristics, and thus removed any differences that could be attributed to these variables. Thus, the finding of the higher likelihood of eclampsia in the Medicaid managed care women relative to women in the private managed care group in California is not likely to be due to differences in characteristics of women in the two payer groups. Because race, which is associated with hypertension, was controlled for in the analysis, the differences in race observed among the payer groups may be ruled out as a factor contributing to the higher likelihood of eclampsia. Although this could be a chance significant finding, it could also have emerged because we did not control for a factor not measured by this study.

It is possible that the increased likelihood of eclampsia was observed in the California Medicaid managed care group because a lower percentage of women received adequate prenatal care compared with the private managed care group in that state. The low prevalence of adequate prenatal care in the Medicaid population relative to other payer groups in California was previously reported.²³ Given a basic tenet of managed care—that is, to promote wellness—it is surprising that the level of prenatal care adequacy was not higher among the Medicaid managed care groups compared with the Medicaid fee-for-service groups. Even in California, where there is more experience relative to Florida with Medicaid managed care, the percentage of women with adequate prenatal care was lower in the Medicaid managed care group than in either the Medicaid fee-for-service or the private managed care groups. A similar finding was observed among Iowa Medicaid recipients.²⁴ However, because prenatal care was controlled for in multiple logistic models used in this study, it is not likely that inadequacies in prenatal care contributed to the higher likelihood of eclampsia occurring during the hospitalization episode.

The lower likelihood of second-degree or higher perineal lacerations in the Medicaid managed care group relative to the private managed care group may be due to individual physician practice styles (eg, use of episiotomies or forceps), practice guidelines of physicians contracted to serve Medicaid managed care groups, or some clinical features of the Medicaid managed care women not controlled for in the analyses.

Although clinical obstetric complications were controlled for in the multivariate analysis, the distribution of high risk in the California sample at first appears counterintuitive. There are two possible explanations for the higher percentage of women classified as high risk in the private managed care group: the length of enrollment in a plan and the intensity of prenatal care monitoring and intervention. Because women in the private managed care group are more likely to have prenatal care and have their conditions more closely monitored and observed, even less severe morbidity may be more frequently documented. In addition, providers in a private managed care plan, at financial risk for hospitalization, may be more aggressive in monitoring (and searching for) potential obstetric complications for the purpose of intervening with the intent of avoiding hospitalization for pregnancy complication, which is a great cost to the health plan. The authors believe that the proportion of mothers determined to have clinical risk factors is a function of the characteristics of the population of women and of how long women are observed in prenatal care and the intensity of prenatal care activities within the health plan.

The lack of differences reported here for a number of the quality indicators that were examined is not likely to be due to a type II error. There was over 95% power in both California and Florida to detect an OR of an adverse outcome of 1.2 based on the lowest prevalence of morbidity outcome in that state. The overall mean length of hospital stay was comparable among the payer groups and thus is not likely to account for the observed differences in the obstetric outcomes differentially occurring because of longer hospitalization time. In addition, if coding errors existed in adverse outcomes, there is no reason to believe that there was systematic difference in coding among the study groups. Moreover, codes from both the birth certificate and the hospital discharge abstract were considered in determining the presence of an adverse outcome.

The adverse outcomes selected for study were not graded with respect to hypothesized level of importance or severity. Future studies may consider the construction of a grading system for the outcomes examined here as well as the possibility of identifying clusters of quality indicators. For example, clusters of adverse outcomes may vary by age or race/ethnicity of the mother. Clustering adverse outcomes into severity levels may obviate some of the statistical problems associated with the use of low incidence adverse outcomes.

The major limitation of this study was its design. Mothers studied were not randomly assigned into the Medicaid managed care or the Medicaid fee-for-service groups. Therefore, selection bias and differences in health status not measured by the variables in these data sets may account for some of the differences seen. We attempted to correct in part for the health status effect through the identification of high-risk status using ICD-9 codes on the hospital discharge abstract and from demographic information from the birth certificate. Another limitation of the study was that characteristics of the health care professional managing the maternal delivery episode were not included in our multivariate models of the likelihood of adverse outcomes. Characteristics of the practitioner, the plan, its financial incentive structures, and profit/not-for-profit status therefore may account for the differences observed. Finally, we did not adjust for length of enrollment in Medicaid or plan organizational form, and these may also account for differences observed or missed.

This study examined the impact of managed care on hospital obstetric care outcomes using a large population of mothers to address the sample size limitation of previous studies on this topic. Medicaid managed care was not found to affect hospital obstetric care adversely, at least for mothers. From a public policy point of view, if managed care can control health care costs and provide at least a comparable level of quality, greater attention can be given to efforts to promote the health of mothers and children. However, the identification of even one indicator of increased risk of an adverse outcome in the Medicaid managed care group relative to the private managed care group suggests that parity in pregnancy outcomes has not yet been achieved. This difference may not necessarily be due to the payer systems, but may be a result of underlying characteristics of the Medicaid population not measured by this study and differences that could not be modified even with changes in payer system.

	Footnotes

 
This work was supported by Health Care Financing Administration grant no. 18-P-90249/5. The statements contained in this article are solely those of the authors and do not necessarily reflect the views or policies of the Health Care Financing Administration, the Agency for Health Care Policy and Research, the State of California, nor the State of Florida.

The authors are very grateful for the programming assistance of Lee D. Thompson, PhD.

PII S0029-7844(99)00534-7

Received September 28, 1998. Received in revised form June 6, 1999. Accepted July 15, 1999.

	References

Top Abstract Materials and Methods Results Discussion References

 
1. United States General Accounting Office. Medicaid: Spending pressures drive states toward program reinvention. GAO/HEHS-95-122. Washington, DC: US Government Printing Office, 1995.

2. United States General Accounting Office. Medicaid: States turn to managed care to improve access and control costs. GAO/HRD-93-46. Washington, DC: US Government Printing Office, 1993.

3. Carey TS, Weis K, Homer C. Prepaid versus traditional Medicaid plans: Lack of effect on pregnancy outcomes and prenatal care. Health Serv Res 191;26:165–81.

4. Krieger JW, Connell FA, LoGerfo JP. Medicaid prenatal care: A comparison of use and outcomes in fee-for-service and managed care. Am J Public Health 1992;82:185–90.[Abstract/Free Full Text]

5. Balaban D, McCall N, Jones-Bauer E. Quality of Medicaid managed care: An evaluation of the Arizona Health Care Cost Containment System. San Francisco, California: Laguna Research Associates. Health Policy Research Series (Discussion paper 94-2), 1994.

6. Stafford RS. Cesarean section use and source of payment: An analysis of California hospital discharge abstracts. Am J Public Health 1990;80:313–5.[Abstract/Free Full Text]

7. Tussing AD, Wojtowycz MA. Health maintenance organizations, independent practice associations, and cesarean section rates. Health Serv Res 1994;29:75–93.[Medline]

8. Oleske DM, Glandon GL, Tancredi DJ, Nassirpour M, Noak JR. Information dissemination and the cesarean birth rate: The Illinois experience. Int J Technol Assess Health Care 1992;8:708–18.[Medline]

9. King DE, Lahiri K. Socioeconomic factors and the odds of vaginal birth after cesarean delivery. JAMA 1994;272:524–9.[Abstract]

10. American College of Obstetricians and Gynecologists. Quality assurance in obstetrics and gynecology. Washington, DC: American College of Obstetricians and Gynecologists, 1989.

11. Joint Commission on Accreditation of Healthcare Organizations. Overview of the indicator monitoring system. Oakbrook Terrace, Illinois: Joint Commission for the Accreditation of Healthcare Organizations, 1992.

12. Jones L, Lo Gerfo J, Shy K, Connell F, Holt V, Parrish K, et al. StORQS: Washington’s statewide obstetrical review and quality system: Overview and provider evaluation. Qual Rev Bull 1993; 110–8.

13. Hordnes K, Bergsjo P. Severe lacerations after childbirth. Acta Obstet Gynecol Scand 1993;72:413–22.[Medline]

14. Welt SI, Cole JS, Myers MS, Sholes DM, Jelovsek FR. Feasibility of postpartum rapid hospital discharge: A study from a community hospital population. Am J Perinatol 1993;10:384–7.[Medline]

15. Philipson EH, Curry SL. Quality assurance: Measuring its effect on a busy obstetric service. Obstet Gynecol 1994;83:131–3.[Abstract/Free Full Text]

16. Kotelchuck M. The Adequacy of Prenatal Care Utilization Index: Its US distribution and association with low birthweight. Am J Public Health 1994;84:1486–9.[Abstract/Free Full Text]

17. Kotelchuck M. An evaluation of the Kessner adequacy of prenatal care index and a proposed adequacy of prenatal care utilization index. Am J Public Health 1994;84:1414–20.[Abstract/Free Full Text]

18. American College of Obstetricians and Gynecologists. Standards for obstetric-gynecologic services. 6th ed. Washington, DC: American College of Obstetricians and Gynecologists, 1985.

19. American Academy of Pediatrics and American College of Obstetricians and Gynecologists. Guidelines for perinatal care. Washington, DC: American Academy of Pediatrics and American College of Obstetricians and Gynecologists, 1988.

20. Oleske DM. Hospital obstetric care: A comparison of quality indicators in Medicaid fee-for-service and Medicaid managed care groups. Final report (HCFA grant no. 18-P-90429/5), 1997.

21. American Hospital Association. 1993 annual survey of hospitals [computer tape]. Chicago, Illinois: American Hospital Association, 1994.

22. Oleske DM, Branca ML, Schmidt JB, Ferguson R, Linn ES. A comparison of capitated and fee-for-service Medicaid reimbursement methods on pregnancy outcomes. Health Serv Res 1998;33: 55–73.

23. Kessler S, Shah R, Smith T, Taylor D. Adequacy of prenatal-care utilization—California, 1989–1994. MMWR Morb Mortal Wkly Rep 1996;45:653–6.[Medline]

24. Schulman ED, Sheriff DJ, Momany ET. Primary care case management and birth outcomes in the Iowa Medicaid program. Am J Public Health 1997;87:80–4.[Abstract/Free Full Text]

This article has been cited by other articles:

				R. H. Miller and H. S. Luft HMO Plan Performance Update: An Analysis Of The Literature, 1997-2001 Health Aff., July 1, 2002; 21(4): 63 - 86. [Abstract] [Full Text] [PDF]

				S. E. Gerber, D. G. Dobrez, and P. P. Budetti Managed Care and Perinatal Regionalization in Washington State Obstet. Gynecol., July 1, 2001; 98(1): 139 - 143. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by OLESKE, D. M. Articles by SMITH, T. Search for Related Content PubMed PubMed Citation Articles by OLESKE, D. M. Articles by SMITH, T.